On 23 July 2026, Cognition announced that it had acquired The Interaction Company of California, the startup behind Poke, in a deal the two parties valued in the ‘low nine figures.’ The strategic logic is straightforward: Cognition acquires Poke not simply to expand its product range, but to transplant a personality into its enterprise coding agent Devin.
Cognition Acquires Poke for Personality, Not Just Code
Poke, which launched in March 2026, is an AI assistant users access through their messaging platform of choice: iMessage, SMS, Telegram, or, in certain markets, WhatsApp. It covers tasks across travel, health, finance, scheduling, and education, with productivity work such as email management and reminders among its most common uses.
What distinguishes Poke is its interaction style. Rather than waiting to be summoned, Dealroom reports that Poke messages users first, incorporating slang and humour rather than adopting a formal tool-like register. Over the past three months, users exchanged more than 100 million messages on the platform, according to Cognition, with hundreds of thousands of people using the service.
Despite that scale, profitability had proved elusive. Marvin von Hagen, co-founder of The Interaction Company, acknowledged in an interview with TechCrunch that running Poke had been expensive, making it difficult to turn a profit.
The acquisition gives Poke access to Cognition’s models and infrastructure. In return, Devin gets something its investors apparently felt it lacked: a warmer presence. ‘You probably prefer it if you have co-workers that have personality, rather than if you have co-workers that are just robots,’ von Hagen told TechCrunch. ‘When you have co-workers that are also software engineers, they can also make a joke. And you’ll find it funny and enjoyable.’
Cognition co-founder Scott Wu framed the rationale in his acquisition post. ‘The Interaction team has built an agent that people love: it’s proactive, it knows you, and it’s fun to talk to. That’s exactly how working with Devin should feel, and now we get to build it together,’ he wrote. ExplainX reports that Wu also described the team as ‘exceptional people who move extremely fast and aim for only the most ambitious products.’
Wu and fellow Cognition co-founder Walden Yan had previously invested in The Interaction Company as angel backers, a detail Wu confirmed in the blog post.
A Second Acquisition Shapes Cognition’s Broader Ambition
The Poke deal is the second major acquisition Cognition has made in quick succession. The company earlier bought Windsurf’s intellectual property, product, and staff following Google’s hire of much of Windsurf’s team. Yahoo Finance reports that Windsurf had reached $82 million in annual recurring revenue at the time, with enterprise ARR doubling quarter-over-quarter and at least 350 enterprise customers alongside hundreds of thousands of daily active users.
That context matters for understanding how Cognition sees itself. The company was one of the first AI startups to launch a fully fledged coding agent, pitching Devin as a junior software engineer capable of handling tasks end-to-end, rather than the augmentation tools offered by rivals such as Cursor and Windsurf. Early reviewers found that Devin made mistakes. Windsurf’s ARR figures suggest the more cautious, developer-assist model attracted significant commercial traction, a gap Cognition is now actively working to close.
Dealroom reports that Cognition frames both Devin and Poke as expressions of a single thesis: always-on cloud agents that work in the background and report back, with Devin targeting enterprise software engineering and Poke handling everyday consumer life.
In June, Poke became the first AI agent approved to run on Apple‘s Messages for Business platform, which gives businesses a standardised channel for customer communications within the Messages app. That relationship is expected to continue.
For now, nothing changes for Poke users through the end of 2026. Next year, the two teams plan to experiment more directly: Poke will draw on Cognition’s SWE-1.7 software engineering model for certain tasks, and the teams will explore how the two products can improve each other. Von Hagen sketched one possibility: ‘I think in the long term Poke can be more reliable with orchestrating all these coding [tasks]…and, at the same time, Devin could become more like Poke. Like right now, Devin can only do one PR (pull request) at a time…I think there’s a lot of value in having Poke orchestrate different Devin sessions.’
A full merger of the two products has not been ruled out. The more immediate test is whether adding conversational warmth to an enterprise coding tool translates into the kind of retention that pure capability has so far struggled to deliver.
